02 dicembre 2014

What and Why

A quick background:

  • Among young adults (15-29), suicide accounts for 8.5% of all deaths: second leading cause of death (after traffic accidents).
  • Among adults (30-49) suicide accounts for 4.1% of all deaths: fifth leading cause of death.

(see the latest "Prevent Suicide" report by WHO)

  • "The debate about the impact of climate change on human health has, very recently, included consideration of mental health" (Berry, Bowen & Kjellstrom, 2009).

  • The mental health vs climate relationship still needs to be studied into details (research gap).

Research question:

How do climate factors affect suicide rate in the Italian context?

  • Is suicide rate lower whereby the average temperature is higher?
  • Is suicide rate higher whereby the average temperature range is higher?
  • Is suicide rate higher whereby precipitations happen more frequently?

Methodology

Case selection:

Previous study: regional data, 20 observations, unsignificant model.

Now, Italy is still our focus: county-level data provide 110 observations!

Our Models:

\[ LogStSuicideRate = \alpha + \beta_{1}LogGDPPerCapita + \beta_{2}LogUnemploymentRate \] \[ + \beta_{3}LogPopulation + \beta_{4}LogGiniIndex + \beta_{5}LogAverageTemperature \] \[ + \beta_{6}LogAverageTemperatureRange + \beta_{7}LogAveragePrecipitations + e \]

\[ LogSuicideRate = \alpha + \beta_{1} + \beta_{2}LogUnemploymentRate \] \[ + \beta_{3}LogPopulation + \beta_{4}LogAverageTemperature \] \[ + \beta_{5}LogAveragePrecipitations + e \]

Findings

Models Comparison
Dependent variable:
St. suicide rate
(1) (2)
GDP per capita 0.699*** 0.529***
(0.211) (0.117)
Unemployment rate 0.086
(0.106)
Population -0.174*** -0.158***
(0.043) (0.039)
Gini index 0.173
(0.487)
Average Temperature -0.072 -0.069
(0.097) (0.092)
Temperature range -0.066
(0.211)
Average Precipitations -0.435*** -0.461***
(0.147) (0.143)
Constant 5.312*** 5.622***
(1.415) (1.123)
Observations 103 103
R2 0.372 0.365
Adjusted R2 0.326 0.339
Residual Std. Error 0.264 (df = 95) 0.261 (df = 98)
F Statistic 8.041*** (df = 7; 95) 14.086*** (df = 4; 98)
Note: p<0.1; p<0.05; p<0.01
Models Comparison
Model1 Model2

Predicted Probabilities

How to cite this model in Zelig: Kosuke Imai, Gary King, and Olivia Lau. 2014. "ls: Least Squares Regression for Continuous Dependent Variables" in Kosuke Imai, Gary King, and Olivia Lau, "Zelig: Everyone's Statistical Software," http://gking.harvard.edu/zelig